Study Examines Prevalence of Schizophrenia Cases Associated with Cannabis Use Disorder After Cannabis Legalization

Image | adobe.stock/freshidea

This article has been updated to include additional commentary on interpreting the findings.

In a recently published study (1), researchers examined the potential connections between cannabis legalization in Ontario, Canada and the prevalence of schizophrenia diagnoses. The study was conducted from January 1, 2006, to December 31, 2022, which covers both the legalization of medical cannabis, which started in 2001 and expanded until full legalization in 2015, and nonmedical cannabis legalization in 2018. The population-based cohort study, “Changes in Incident Schizophrenia Diagnoses Associated With Cannabis Use Disorder After Cannabis Legalization,” was published in February 2025 in JAMA Network Open (1). It was funded by Canadian Institutes of Health Research and by ICES (formerly known as the Institute for Clinical Evaluative Sciences).

The study noted prior research on associations of cannabis use, especially cannabis with high THC content, and the development or earlier onset of psychosis and schizophrenia, though evidence on cannabis legalization and psychosis is deficient. The goal in this study was to measure the population-attributable risk fraction (PARF) for cannabis use disorder (CUD) associated with schizophrenia, in other words, the number of cases of schizophrenia that could have been prevented if cannabis use severe enough to require emergency or hospital care was eliminated.

This study analyzed data from 13,588,681 people aged 14 to 65 years without a history of schizophrenia, noting diagnoses of CUD diagnoses in emergency rooms or hospital settings.

Some of the findings included:

• Out of the total population, 118,650 individuals had CUD
• 91,106 developed schizophrenia (0.6% of the population without CUD and 8.9% of the population with CUD)
• From the prelegalization period to the legalization period, the PARF for CUD associated with schizophrenia increased from 3.7% to 10.3%
• Compared with the PARF for CUD associated with schizophrenia, which increased steadily even after cannabis policy changes, the PARF for CUD associated with psychosis not otherwise specified accelerated after the liberalization of medical cannabis policy

Limitations of the study were noted. These included a lack of access to confounders (including individual-level income, educational attainment, family history of mental health disorders, and genetics), which could affect the validity of the PARF estimators. Other limitations included the exposure definition of CUD and the impacts of COVID-19.

Ultimately, the researchers concluded: “the proportion of incident cases of schizophrenia associated with CUD almost tripled during a period encompassing ongoing liberalization of medical and nonmedical cannabis. Although the proportion of cases of schizophrenia associated with CUD increased fairly linearly over time, incident cases of psychosis NOS and the proportion associated with CUD accelerated after cannabis liberalization.”

More research was called for in order to understand cannabis legalization policy’s long-term associations with the occurrence of psychotic disorders. Additionally, it was suggested that the association between cannabis use disorders and schizophrenia be taken into consideration in the legalization of cannabis.

Invited Editorial Commentary on this Study

Included in the study is editorial commentary (2) from Jodi M. Gilman, PhD, Center for Addiction Medicine, Department of Psychiatry, Massachusetts General Hospital and Harvard Medical School. Dr. Gilman summarized the study and offers insight on the major findings, explaining that the first main finding provides more evidence for the association between cannabis use and an increased risk for psychosis, especially in young adults.

She comments on the second main finding: “The second finding of this study is that the PARF for CUD associated with psychosis NOS, but not schizophrenia, accelerated further with the legalization of cannabis for medical but not recreational purposes. These mixed findings should not make us doubt the association between cannabis use and psychosis; rather, they highlight the methodological challenges of linking the biological causality of a complex psychiatric illness, such as schizophrenia, with a cannabis policy change.”

Dr. Gilman also addresses three challenges of using large-scale electronic medical record studies when assessing mental health consequences of cannabis policy, specifically, establishing concrete time points of legalization, establishing the start of cannabis use and the onset of psychosis, and establishing data on THC potency in cannabis consumed. She concludes: “It is important not to misinterpret findings suggesting no significant acceleration in adverse outcomes after cannabis legalization as a statement that policy does not matter. Policy does matter; allowing unfettered commercial markets to exist exposes more individuals to a greater variety of readily available, high-potency cannabis products.”

Correlation vs Causation; Interpreting the Findings

Ruth Fisher, PhD, contributor to Cannabis Science and Technology offers additional insights and commentary on interpreting the results of the study. First, she notes that the populations of CUD and schizophrenia are very small: of the people with schizophrenia, 11.6% had CUD, while 88.4% did not, and of the people with CUD, 8.9% had schizophrenia, while 0.6% did not. “There’s a nontrivial coincidence – this only establishes correlation, not causation,” she explains.

Fisher also comments on the increase in accessibility of cannabis from 2006 to 2022 and the incidence of schizophrenia, noting that the incidence rate of schizophrenia decreased between 2006 to 2022, and the incidence of schizophrenia stayed the same between pre- and post-legalization. “The researchers simply glossed over this fact without providing any explanation and simply shifted the focus to cases of schizophrenia with CUD,” Fisher states. “It seems utterly plausible to me that people with schizophrenia don’t feel well, so they try to medicate. If cannabis became easier to access over the period, then it makes sense they’d increasingly use cannabis to self-medicate.”

She also points out that the analysis does not account for any other factors happening in the sociopolitical environment during this period, such as changes in polices regarding diagnosis or treating people with substance use or mental health disorders.

Fisher concludes: “In short, the analysis presented looks at several different co-occurring factors – cannabis legalization, CUD diagnosis, schizophrenia diagnosis, psychosis – and tries to present correlations as if they’re causations, without accounting for other more plausible explanations for the patterns found. Notably, they fail to recognize that the incidence of CUD and schizophrenia are both extremely small, and they seem to gloss over the fact that the incidence of schizophrenia was stable, if not decreasing over the period while cannabis accessibility vastly increased. The authors also fail to recognize the tremendous harm reduction impact of cannabis on society – it’s possible that without the availability of cannabis, people would have fared even worse.”

References

1. Myran, DT.; Pugliese, M.; Harrison, LD.; Solmi, M.; Anderson, KK.; Fiedorowicz, JG.; Finkelstein, Y.; Manuel, D.; Taljaard, M.; Webber, C.; Tanuseputro, P. Changes in Incident Schizophrenia Diagnoses Associated With Cannabis Use Disorder After Cannabis Legalization. JAMA Netw Open. 2025. DOI: 10.1001/jamanetworkopen.2024.57868
2. Gilman, Jodi. Association of Cannabis Legalization With Prevalence of Schizophrenia—Challenges of Attributing Biological Causality to Policy Change https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2829844